结合SURF与聚类分析方法实现运动目标的快速跟踪  

Moving Target Fast Tracking Using SURF and Cluster Analysis Method

作　　者：李英[1] 李静宇[2,3] 徐正平[2] 

机构地区：[1]长春理工大学光电工程学院,吉林长春130022 [2]中国科学院长春光学精密机械与物理研究所,吉林长春130033 [3]中国科学院研究生院,北京100039

出　　处：《液晶与显示》2011年第4期544-550,共7页Chinese Journal of Liquid Crystals and Displays

摘　　要：为了解决运动目标快速跟踪过程的实时性与稳定跟踪问题,提出了结合SURF(Speed Up Robust Features)与K-means聚类分析的运动目标快速跟踪算法(SURF-KMs),对图像的局部多尺度特征提取与描述进行了研究。首先,使用SURF算法在跟踪窗口内提取特征点,生成并匹配特征矢量。然后,利用K-means算法估计目标特征点的质心位置,确定其聚集范围,实时更新窗口尺寸和位置。最后,建立目标模板更新策略,当目标发生形态变化而无遮挡时,更新目标模板。实验结果表明,当目标发生大角度旋转和快速缩放,同时发生颜色变化时,所提出的SURF-KMs算法仍能够实现稳定的跟踪,且满足运动目标实时跟踪的稳定可靠、精确度高、抗干扰能力强等指标要求。In order to design a moving target fast tracking system with respect to a time-limited and stable tracking process, specially when the shape of moving objective or its environment condition change, a new approach named SURF-KMs which combines the advantage of SURF algorithm with a cluster analysis of K-means method is proposed. First, based on SURF algorithm, interest points and vectors are presented. Second, using K-means method, we can estimate the target-center, determine target^s cluster scope and update the tracking window online. Finally, a self-adapting updating strategy for matching template is also pro- posed in order to track moving target automatically. Experimental results indicate that SURF-KMs is mostly able to achieve a stable tracking while the monitored target rotating, scale changing, and also the environment illumination glittering. Moreover, it can satisfy the system requirements of tracking stability, higher precision and anti-jamming.

关 键 词：SURF 聚类分析 运动目标 快速跟踪 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]